Is There a Critical Age or Period for Learning Languages? | Dr. Erich Jarvis & Dr. Andrew Huberman
00:00:00.000 | And how similar or different are the brain areas
00:00:05.000 | controlling speech and language in say a songbird
00:00:08.800 | and a young human child?
00:00:10.800 | - Yeah, so going back to the 1950s
00:00:14.440 | or even a little earlier and Peter Mahler and others
00:00:17.800 | who got involved in neuroethology,
00:00:19.840 | the study of neurobiology of behavior in a natural way,
00:00:24.160 | right, you know, they start to find that behaviorally,
00:00:29.240 | there are these species of birds like songbirds and parrots
00:00:31.960 | and now we also know hummingbirds, just three of them
00:00:34.440 | out of the 40 something bird groups
00:00:36.080 | out there on the planet, orders,
00:00:38.360 | that they can imitate sounds like we do.
00:00:41.040 | And so that was a similarity.
00:00:42.720 | In other words, they had this kind of behavior
00:00:44.560 | that's more similar to us than chimpanzees have with us
00:00:47.660 | or than chickens have with them, right,
00:00:50.060 | their closer relatives.
00:00:51.560 | And then they discovered even more similarities,
00:00:54.040 | these critical periods that if you remove a child,
00:00:58.540 | you know, this unfortunately happens where a child is feral
00:01:01.720 | and is not raised with human
00:01:03.080 | and goes through their puberty phase of growth,
00:01:06.360 | becomes hard for them to learn a language as an adult.
00:01:09.600 | So there's this critical period where you learn best.
00:01:12.400 | And even later on when you're in regular society,
00:01:14.760 | it's hard to learn.
00:01:15.900 | Well, the birds undergo the same thing.
00:01:18.760 | And then it was discovered that if they become deaf,
00:01:22.360 | we humans become deaf, our speech starts to deteriorate
00:01:25.880 | without any kind of therapy.
00:01:28.380 | If a non-human primate or, you know,
00:01:31.600 | or let's say a chicken becomes deaf,
00:01:34.900 | their vocalizations don't deteriorate,
00:01:36.800 | very little, at least.
00:01:38.800 | Well, this happens in the vocal learning birds.
00:01:41.280 | So there were all these behavioral parallels
00:01:43.200 | that came along with the package.
00:01:44.920 | And then people looked into the brain,
00:01:47.320 | Fernando Nadava, my former PhD advisor,
00:01:49.920 | and began to discover the area X you talked about,
00:01:53.720 | the robust nucleus of the archipelagium.
00:01:56.280 | And these brain pathways were not found
00:02:00.360 | in the species who couldn't imitate.
00:02:01.720 | So there was a parallel here.
00:02:03.520 | And then jumping many years later, you know,
00:02:06.800 | I started to dig down into these brain circuits
00:02:09.760 | to discover that these brain circuits
00:02:12.160 | have parallel functions with the brain circuits for humans,
00:02:15.160 | even though they're by a different name,
00:02:16.600 | like Broca's and laryngeal motor cortex.
00:02:19.280 | And most recently we discovered not only
00:02:22.340 | the actual circuitry and the connectivity are similar,
00:02:24.880 | but the underlying genes that are expressed
00:02:27.560 | in these brain regions in a specialized way,
00:02:30.200 | different from the rest of the brain are also similar
00:02:32.800 | between humans and songbirds and parrots.
00:02:35.200 | So all the way down to the genes,
00:02:36.520 | and now we're finding the specific mutations
00:02:39.640 | are also similar, not always identical, but similar,
00:02:43.240 | which indicates remarkable convergence
00:02:45.400 | for a so-called complex behavior
00:02:47.760 | in species separated by 300 million years
00:02:49.920 | from a common ancestor.
00:02:51.240 | And not only that, we are discovering
00:02:53.780 | that mutations in these genes
00:02:56.720 | that cause speech deficits in humans, like in FOXP2,
00:03:00.520 | if you put those same mutations or similar type of deficits
00:03:05.280 | in these vocal learning birds, you get similar deficits.
00:03:08.120 | So convergence of the behavior is associated
00:03:11.160 | with similar genetic disorders of the behavior.
00:03:14.000 | - Incredible.
00:03:15.280 | I have to ask, do hummingbirds sing or do they hum?
00:03:19.400 | - Hummingbirds hum with their wings
00:03:21.320 | and sing with their syrinx.
00:03:23.160 | - In a coordinated way?
00:03:24.540 | - In a coordinated way.
00:03:25.980 | There's some species of hummingbirds
00:03:28.820 | that actually will,
00:03:30.580 | Doug Oshawa showed this,
00:03:32.860 | that will flap their wings
00:03:35.420 | and create a slapping sound with their wings
00:03:38.400 | that's in unison with their song.
00:03:40.700 | And you would not know it,
00:03:42.620 | but it sounds like a particular syllable in their songs,
00:03:45.960 | even though it's their wings
00:03:48.740 | and their voice at the same time.
00:03:50.500 | - Hummingbirds are clapping to their song.
00:03:52.900 | - Clapping, they're snapping their wings together
00:03:56.660 | in unison with the song to make it like,
00:03:59.260 | if I'm going,
00:04:00.100 | ♪ Ba-da-da-da-da, ba-da ♪
00:04:02.460 | You know, and I banged on the table.
00:04:04.160 | Except they make it almost sound like their voice
00:04:06.500 | with their wings.
00:04:07.440 | - Incredible.
00:04:09.420 | - Yes.
00:04:10.260 | - I'm-
00:04:11.080 | - And they got some of the smallest brains around.
00:04:11.920 | - I guess as a kid, you would say mind blown.
00:04:13.740 | - Yes, yes.
00:04:14.580 | - Incredible.
00:04:15.400 | - Yes.
00:04:16.240 | - Incredible, I love hummingbirds.
00:04:17.100 | And I always feel like it's such a special thing
00:04:19.500 | to get a moment to see one because they move around so fast
00:04:21.820 | and they flit away so fast in these ballistic trajectories
00:04:25.500 | that when you get to see one stationary for a moment,
00:04:29.100 | or even just hovering there,
00:04:31.220 | you feel like you're extracting so much
00:04:32.780 | from their little microcosm of life.
00:04:35.540 | But now I realize they're playing music, essentially.
00:04:38.100 | - Right, exactly.
00:04:38.940 | And what's amazing about hummingbirds,
00:04:41.420 | and we're gonna say vocal learning species in general,
00:04:44.740 | is that for whatever reason,
00:04:46.620 | they seem to evolve multiple complex traits.
00:04:50.260 | You know, this idea that evolving language,
00:04:53.260 | spoken language in particular,
00:04:55.240 | comes along with a set of specializations.
00:04:58.140 | - Incredible.
00:04:59.700 | When I was coming up in neuroscience,
00:05:02.020 | I learned that, I think it was the work of Peter Marler,
00:05:05.960 | that young birds learn, songbirds,
00:05:09.660 | learn their tutor's song and learn it quite well,
00:05:14.660 | but that they could learn the song of another tutor.
00:05:17.900 | In other words, they could learn a different,
00:05:19.580 | and for the listeners, I'm doing air quotes here,
00:05:21.260 | a different language, a different bird song,
00:05:24.020 | different than their own species song,
00:05:25.980 | but never as well as they could learn
00:05:28.060 | their own natural genetically linked song.
00:05:32.860 | - Yes.
00:05:33.680 | - Genetically linked, meaning that it would be like
00:05:35.540 | me being raised in a different culture,
00:05:37.260 | and that I would learn the other language,
00:05:41.300 | but not as well as I would have learned English.
00:05:44.140 | This is the idea.
00:05:45.140 | - Yes.
00:05:45.980 | - Is that true?
00:05:46.800 | - That is true, yes.
00:05:47.760 | And that's what I learned growing up as well,
00:05:49.540 | and talked to Peter Marler himself about before he passed.
00:05:52.880 | Yeah, he used to call it the innate predisposition to learn.
00:05:57.880 | All right, so, which would be kind of the equivalent
00:06:02.240 | in the linguistic community of universal grammar.
00:06:04.780 | There is something genetically influencing
00:06:09.580 | our vocal communication on top of what we learn culturally.
00:06:13.860 | And so there's this balance between
00:06:15.980 | the genetic control of speech, or a song in these birds,
00:06:19.780 | and the learned cultural control.
00:06:22.860 | And so, yes, if you were to take,
00:06:25.620 | I mean, in this case, we actually tried this
00:06:29.980 | at Rockefeller later on, take a zebra finch
00:06:33.100 | and raise it with a canary, it would sing a song
00:06:36.640 | that was sort of like a hybrid in between,
00:06:38.260 | we call it a caninch, right?
00:06:40.240 | (laughing)
00:06:41.820 | And vice versa for the canary,
00:06:43.640 | because there's something different
00:06:44.700 | about their vocal musculature or the circuitry in the brain.
00:06:48.940 | And with a zebra finch, even with a closely related species,
00:06:52.140 | if you would take a zebra finch, a young animal,
00:06:55.700 | and in one cage next to it,
00:06:57.220 | place its own species, adult male, right?
00:07:00.500 | And in the other cage, place a Bengalis finch next to it,
00:07:03.860 | it would preferably learn the song
00:07:06.060 | from its own species neighbor.
00:07:08.780 | But if you remove its neighbor,
00:07:10.540 | it would learn that Bengalis finch very well.
00:07:13.300 | - Fantastic. - So there's,
00:07:14.780 | it has something to do with also the social bonding
00:07:17.380 | with your own species.
00:07:18.740 | - Incredible.
00:07:19.960 | That raises a question that I,
00:07:21.820 | based on something I also heard,
00:07:23.220 | but I don't have any scientific,
00:07:25.620 | peer-reviewed publication to point to,
00:07:27.340 | which is this idea of pigeon, not the bird,
00:07:29.860 | but this idea of when multiple cultures and languages
00:07:34.140 | converge in a given geographic area,
00:07:36.040 | that the children of all the different native languages
00:07:38.780 | will come up with their own language.
00:07:41.580 | I think this was in island culture,
00:07:43.380 | maybe in Hawaii, called Pidgin,
00:07:45.020 | which is sort of a hybrid of the various languages
00:07:47.940 | that their parents speak at home,
00:07:49.740 | and that they themselves speak,
00:07:51.740 | and that somehow Pidgin, again, not the bird,
00:07:54.980 | but a language called Pidgin for reasons I don't know,
00:07:57.780 | harbors certain basic elements of all language.
00:08:03.020 | Is that true?
00:08:04.180 | Is that not true?
00:08:05.060 | - I haven't studied enough myself
00:08:07.780 | in terms of Pidgin specifically,
00:08:09.900 | but in terms of cultural evolution of language
00:08:12.540 | and hybridization between different cultures and so forth,
00:08:15.940 | even amongst birds with different dialects,
00:08:18.940 | and you bring them together,
00:08:20.460 | what is going on here is cultural evolution
00:08:26.340 | remarkably tracks genetic evolution.
00:08:29.540 | So if you bring people
00:08:31.620 | from two separate populations together
00:08:33.460 | that have been in their separate populations,
00:08:35.740 | evolutionarily at least, for hundreds of generations,
00:08:39.740 | so someone speaking Chinese, someone speaking English,
00:08:42.940 | and that child then is learning from both of them,
00:08:46.620 | yes, that child's gonna be able to pick up
00:08:49.060 | and merge phonemes and words together
00:08:54.060 | in a way that an adult wouldn't,
00:08:56.940 | because why they're experiencing both languages
00:09:00.100 | at the same time during their critical period years
00:09:04.420 | in a way that adults would not be able to experience.
00:09:08.060 | And so you get a hybrid.
00:09:09.940 | And the lowest common denominator
00:09:11.980 | is gonna be what they share.
00:09:13.740 | And so the phonemes that they've retained
00:09:16.180 | in each of their languages
00:09:18.900 | is what's gonna be, I imagine, used the most.
00:09:22.180 | (upbeat music)
00:09:24.780 | (upbeat music)
00:09:27.360 | (upbeat music)